Research map / Probabilistic and scientific modeling / Molecules and proteins
CryoKRAQEN: Kernel-Regularized Annealing for Quantized Embedding Networks in Cryo-EM Heterogeneous Reconstruction
Wenyuan Gao, Yutan Wu, Xuming He
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use triplane representation instead of voxel-based implicit field (ablation) | CryoKRAQEN · CryoBench IgG-1D (masked) | Per-Image AUC-FSC (mean): 0.324 -> 0.375 | Improved |
| use triplane representation instead of Fourier positional encoding (ablation) | CryoKRAQEN · CryoBench IgG-1D (masked) | Per-Image AUC-FSC (mean): 0.368 -> 0.375 | Improved |
| use Epanechnikov kernel instead of Gaussian kernel (ablation) | CryoKRAQEN · CryoBench IgG-1D (masked) | Per-Image AUC-FSC (mean): 0.353 -> 0.375 | Improved |
| add kernel-guided similarity (ablation) | CryoKRAQEN · CryoBench IgG-1D (masked) | Per-Image AUC-FSC (mean): 0.336 -> 0.375 | Improved |
| add cosine temperature annealing (vs low initial temperature without annealing) (ablation) | CryoKRAQEN · CryoBench IgG-1D (masked) | Per-Image AUC-FSC (mean): 0.309 -> 0.375 | Improved |
| add cosine temperature annealing (vs high initial temperature without annealing) (ablation) | CryoKRAQEN · CryoBench IgG-1D (masked) | Per-Image AUC-FSC (mean): 0.339 -> 0.375 | Improved |
| add triplet loss regularization on quantized embeddings (ablation) | CryoKRAQEN · CryoBench IgG-1D (masked) | Per-Image AUC-FSC (mean): 0.295 -> 0.375 | Improved |
| CryoKRAQEN versus RECOVAR on IgG-1D | CryoKRAQEN · CryoBench IgG-1D | Per-Image AUC-FSC (mean): 0.386 -> 0.375 | Hurt |
| CryoKRAQEN versus CryoDRGN on Tomotwin-100 compositional dataset | CryoKRAQEN · CryoBench Tomotwin-100 | Per-Image AUC-FSC (mean): 0.316 -> 0.335 | Improved |
| CryoKRAQEN versus RECOVAR on Spike-MD | CryoKRAQEN · CryoBench Spike-MD | Per-Image AUC-FSC (mean): 0.362 -> 0.328 | Hurt |
Closest papers by meaning
- Mixture of neural fields for heterogeneous reconstruction in cryo-EM (NeurIPS 2024 · similarity 0.91)
- cryoSPHERE: Single-Particle HEterogeneous REconstruction from cryo EM (ICLR 2025 · similarity 0.90)
- CryoBench: Diverse and challenging datasets for the heterogeneity problem in cryo-EM (NeurIPS 2024 · similarity 0.90)
- CryoSPIN: Improving Ab-Initio Cryo-EM Reconstruction with Semi-Amortized Pose Inference (NeurIPS 2024 · similarity 0.89)
- Reconstructing Heterogeneous Biomolecules via Hierarchical Gaussian Mixtures and Part Discovery (NeurIPS 2025 · similarity 0.89)
- Multiscale guidance of protein structure prediction with heterogeneous cryo-EM data (NeurIPS 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
